Security industry experts explore the challenges of adoption, the importance of measurable customer outcomes and what it takes to prepare for an AI-driven future.
Artificial intelligence (AI), despite being in nearly every conversation you’ll hear about security, is not yet a foundational part of the industry’s technology strategy. Some manufacturers have been using a breadth of AI solutions in their products for years. Many security dealers and integrators are now familiar with agentic AI, a step above ChatGPT, Gemini, Claude, and similar now-everyday tools. But the amorphic term ‘AI’ is still widely considered a buzzword, synonymous with ‘something inevitable designed to streamline processes.’
Since the industry is not quite there yet, this article will explore where security actually stands with AI, and what security leaders are doing to lead with outcomes instead of solutions, choose the appropriate solutions and prepare for an AI-driven future. Fred Bohr, channel account manager, Salient Systems, Austin, Texas, agrees, “We start with the questions the customer needs answered, not with how many cameras they need. That question decides everything else. For general video intelligence, we recommend building it into the platform first. Third-party analytics belong where they’re truly specialized, such as license plate recognition or multi-factor facial recognition. … We also tell customers they don’t need to overhaul everything at once. Start small and grow it in a way that benefits the organization.”
Cost is another issue when it comes to AI, whether it’s being deployed in security solutions or used for standard business processes. “Two things belong in that calculation, and most companies only count one,” Withers says. “The first is the technology: license, integration and the staff time to run it. The second is the people: training, workflow redesign, governance and the adoption work that determines whether anyone uses the thing. The return on the technology is capped by the investment in the people.”
According to Bohr, the biggest hidden factor on the cost side is architecture. “Running analytics as a separate layer means two event stores, two data schemas and two support paths,” he says. “It also means an integration that can break when a vendor changes something. Hardware is the next factor. If a capability requires replacing cameras, it becomes a capital project that may wait years for approval, so we favor analytics that work with existing cameras. Lifecycle is the third. Long gaps between upgrades keep new capabilities and integrations out of reach, and subscription licensing can move customers to a continuous lifecycle without those gaps. On the return side, the strongest cases tie AI to specific outcomes, such as cutting investigation time from 45 minutes to under five, or delivering weekly occupancy reports to facilities. Outcomes like these bring IT, legal and operations into the business case, not just security.”
Read the full article in SDMMAG.com
